In human testing of standard cigarettes and potential reduced exposure products (PREPs), puff topography is typically measured and programmed into smoking machines (puff replicators) to mimic human smoking. The resulting smoke emissions are tested for relative levels of harmful constituents. The puff topography data that is used is summary data;that is, the total puff volume is divided by the number of puffs to yield an average puff volume. The puff replicator is then programmed to reproduce the average puff, with steady, periodic parabolic waveform, and a fixed puff frequency, volume, and duration. Because humans smoke differently as the rod is consumed, and combustion chemistry is highly non-linear, representing smoking behavior with a smoothed periodic waveform may result in a tobacco smoke aerosol with different chemical composition and physical properties than that actually generated by the smoker. It may also underestimate the smokers'exposure to harmful smoke constituents. This is a proposal to test if replicating the true puff profile will result in differences in levels of select volatile organic and particle-bound semi-volatile smoke constituents in mainstream smoke, compared to machine smoking using only the standard smoothed puff data. We will also determine if replicating the true puff profile results in differences in the particle size distribution of mainstream smoke compared to machine smoking using only the standard smoothed puff data. A sample of previously acquired human puff profiles from an existing cohort will be used to replicate both the true puff profile and the smoothed puff profile using a smoking machine, and the smoke emissions generated by each protocol will be compared for levels of select toxic components of mainstream smoke, and the distribution of fine and ultra-fine particles. This study will help to optimize standards for testing new and conventional smoked tobacco products in order to better characterize human exposure to harmful components of tobacco smoke. PUBLIC HEALTH RELEVANCE: In human testing of new combustible tobacco products, puff topography is typically measured and programmed into smoking machines (puff replicators) to mimic human smoking and the resulting smoke emissions are tested for relative levels of harmful constituents. Using only the summary puff data, with steady, periodic parabolic waveform, and a fixed puff frequency, volume, and duration, may underestimate actual exposure to smoke toxins. This study is the first step in determining if the public health research community should replace the currently used equipment and procedures and begin replicating the true puff profile in the testing of new products to more accurately reflect human exposure.